Automatic Creation of Mappings between Classification Systems for Bibliographic Data

Abstract

Classification systems are an important means to provide topic based access to large collections. They are utilized by a number of approaches for faceted browsing, graphical search support and lately also for collection visualisation and analysis. Most of these approaches have been developed with a specific classification system in mind and often exploit some of the inherent characteristics of the system. Collections that are indexed using local or special classification systems cannot benefit from the vast majority of innovative applications developed for the more commonly used classification systems. One way to alleviate this problem is the use of mappings between classification systems. Traditionally, these mappings have been created in a manual and time consuming process involving subject specialists.In this paper, we discuss another approach to automatically create mappings between classification systems. The approach consists of three steps: First, bibliographic data from diverse sources that contain items classified by the required classification systems is aggregated in a single database. Next, a clustering algorithm is used to group individual issues and editions of the same work. The basic idea is that for classification purposes, there is no significant difference across editions and indexing information can thus be consolidated within the clusters. Finally, the clusters containing information from both required systems are added up to create a cooccurrence table. This information can be used to describe correlations between individual classes of the two classification systems and forms the basis of a full mapping between the two systems. First results from an application of this approach to data from German union catalogues and comparing the derived mappings to manually created ones are quite promising and show the potential of this idea.